Can anyone help explain why the following statistics are either Sufficient or not Sufficient?

Mode, Mean, Median, Standard Deviation, Skewness, Kurtosis, Range, IQR

  • $\begingroup$ Perhaps stats.stackexchange.com would be a better venue for this sort of question. $\endgroup$ – Jebruho Nov 4 '12 at 22:57
  • 2
    $\begingroup$ Every one of them is sufficient for some family of distributions. Sufficiency is relative to a set of probability distributions. This is explained in this article: jstor.org/stable/2683116 $\endgroup$ – Michael Hardy Nov 4 '12 at 23:38

A statistic is not 'sufficient' on its own. Sufficiency is with respect to a distribution (or rather a family).

For example, Wikipedia's article on sufficient statistics has

A statistic $T(X)$ is sufficient for underlying parameter $θ$ precisely if the conditional probability distribution of the data $X$, given the statistic $T(X)$, does not depend on the parameter $θ$

So a statistic may be sufficient for some parameter for one distribution but not sufficient for another parameter of another distribution. I can think of a number of distributions in which the sample mean is sufficient for some statistic ... and vastly more where it isn't.

You don't specify either parameters nor distributions in your question, so we can't really say whether any of them are sufficient. It depends on the situation.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy